updated get transcript

Author: Kamalabot

movie_py_explored/get_script.py

@@ -1,24 +1,28 @@
 from moviepy import VideoFileClip, TextClip, CompositeVideoClip
 from faster_whisper import WhisperModel
+from tqdm import tqdm 
 
 # Load video and extract audio
 video = VideoFileClip("./sales_training_awareness_src.mp4")
-video.audio.write_audiofile("sales_training_awareness.wav") # type: ignore
+video.audio.write_audiofile("auto_edit.wav") # type: ignore
 
 # Transcribe with faster-whisper
 model = WhisperModel("small")  # You can use 'small' or 'medium' if needed
-segments, _ = model.transcribe("sales_training_awareness.wav")
-# print(f"Segments are: {segments}")
+# model_v2 = WhisperModel("large-v2") # when using tamil 
+segments, _ = model.transcribe("auto_edit.wav")
+# when using tamil
+# segments, _ = model_v2.transcribe("/content/training_tamil.wav", language="ta")
+
+# segments are lazy generators, so executing below line will take long time
+# Avoid background music, poor mics, or mixed language
+segments = list(tqdm(segments))  # Shows a progress bar
 
 # Build typewriter subtitles for all segments
 subtitle_clips = []
 with open("subtitle.srt", 'w') as fsrt:
     for idx, segment in enumerate(segments):
         fsrt.write(f"{segment.text} : {segment.start}, {segment.end - segment.start}")
         subtitle_clips.append([segment.text, segment.start, segment.end - segment.start])
-        if idx % 5 == 0:
-            print(f"At idx: {idx}")
-            break
 
 
 print("Writing Extracted Subtitle:")